Trigger: Ted_Bear
Each clip was captioned, referring to the character as Ted_Bear.
Description
FAQ
Comments (5)
dude this is amazing! great job. curious how you trained it? sample caption? works so well.
Grok captioned everything for me. lol
"Ted_Bear. A living anthropomorphic teddy bear named Ted_Bear stands on a dark red carpet in a dimly lit indoor hallway, captured from a low to eye-level camera angle at close-medium distance with a slight downward tilt that emphasizes his small stature next to the left leg of a person wearing casual blue denim jeans and white sneakers, his body oriented toward the camera while his head tilts upward with a concerned and slightly sad facial expression marked by furrowed fuzzy brows, wide brown glass-like eyes, and a downturned mouth on his white fabric snout. Ted_Bear has no clothing whatsoever, revealing his classic shaggy light brown plush fur texture across his rounded torso, stubby arms held slightly outward with palms facing down, and short legs planted firmly on the floor. The background features a white paneled door and dark wooden threshold under warm, low-key side lighting that creates soft shadows and highlights the realistic fabric details of his fur. The shot appears static with no visible camera movement, maintaining a cinematic realistic style consistent with the live-action Ted movie and TV series, and no watermarks, text, or logos are present in the frame.. Audio: "Well, yeah, but I'm assuming not happily as she wouldn't have stuck a tongue down my throat."
@anonymous_time_420 whoa is that all one caption? How many steps? It trained pretty well with so many tokens.
@kronos1959777 3000 steps
Yay Ted! I love it! Great work!